Abstract
Biologists usually pursue the adaptationist paradigm in trying to explain the functional significance of animal coloration. Here I collate instances in which coloration may be a poor match in the context of background matching, Batesian mimicry, aposematism, and colour polymorphisms. This can occur because of trade-offs with other functions, relaxed selection from predation, or colour trait neutrality. Also, mechanistic, pleiotropic and chance genetic effects can all result in a poor match to the background environment or to signaling efficiently. While biologists implicitly recognise these constraints placed on adaptive coloration, they rarely explicitly acknowledge the heterodox notion that coloration might be under weak selection or no selection at all. Unfortunately, it is difficult to show this definitively, as illustrated in an investigation into the function of colour polymorphisms in coconut crabs.